High-Precision and High-Robustness Global Localization for Rail Robots
TANG Hengbo1, CHEN Weidong1, WANG Jingchuan1, LIU Shuai2, LI Guobo2, ZHAO Hongdan2
1. Key Laboratory of System Control and Information Processing, Ministry of Education, Department of Automation, Shanghai Jiao Tong University, Shanghai 200240, China;
2. Liaoyuan Power Co. of Jilin Electric Power Co., LTD, Liaoyuan 136200, China
The localization precision and reliability of a rail robot system will be negatively influenced by environmental factors like electromagnetic interference, unstable temperature and illumination. For the problems above, a global localization system for rail robots based on the fusion of odometry and landmark pins is designed. An optimal distribution method of landmark pins is also proposed to improve localization robustness by expanding the distribution difference of landmark pins in each segment. The simulation of wheel spin and landmark pin failure, together with the application results of the substation inspection robot, indicate that the localization precision is at millimeter scale while the system is capable of global localization and can deal with environmental interference.
 Keller Heirich O, Robertson P, Garcia A C, et al. Probabilistic localization method for trains[C]//IEEE Intelligent Vehicles Symposium. Piscataway, USA: IEEE, 2012: 482-487. 刘进,吴汶麒.轨道交通列车定位技术[J].城市轨道交通研究,2001,4(1):30-34. Liu J, Wu W Q. Train positioning technology of railway and mass transit[J]// Urban Mass Transit, 2011, 4(1): 30-34. Acharya A, Sadhu S, Ghoshal T K. State inequality constraint based method for rail navigation[C]//Annual IEEE India Conference: Green Energy, Computing and Communication. Piscataway, USA: IEEE, 2010: 1-6. Heirich O, Robertson P, Garcia A C, et al. Bayesian train localization method extended by 3D geometric railway track observations from inertial sensors[C]//15th International Conference on Information Fusion. Piscataway, USA: IEEE, 2012: 416-423. Hensel S, Hasberg C. Probabilistic landmark based localization of rail vehicles in topological maps[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems. Piscataway, USA: IEEE, 2010: 4824-4829. Hensel S, Hasberg C, Stiller C. Probabilistic rail vehicle localization with eddy current sensors in topological maps[J]. IEEE Transactions on Intelligent Transportation Systems, 2011, 12(4): 1525-1536.  Wohlfeil J. Vision based rail track and switch recognition for self-localization of trains in a rail network[C]//IEEE Intelligent Vehicles Symposium. Piscataway, USA: IEEE, 2011: 1025-1030. 徐海贵,王春香,杨汝清,等.磁传感系统在室外移动机器人导航中的研究[J].机器人,2007,29(1):61-66.Xu H G, Wang C X, Yang R Q, et al. A magnetic sensing system for outdoor mobile robot navigation[J]. Robot, 2007, 29(1): 61-66. 王景川,方毅,陈卫东.移动机器人定位的自适应功率调节射频识别系统[J].上海交通大学学报,2012,46(2):207-212.Wang J C, Fang Y, Chen W D. Mobile robot self-localization based on RFID system with adaptive power control[J]. Journal of Shanghai Jiao Tong University, 2012, 46(2): 207-212. 许俊勇,王景川,陈卫东.基于全景视觉的移动机器人同步定位与地图创建研究[J].机器人,2008,30(4):289-297.Xu J Y, Wang J C, Chen W D. Omni-vision-based simultaneous localization and mapping of mobile robots[J]. Robot, 2008, 30(4): 289-297. 刑文训,谢金星.现代优化计算方法[M].北京:清华大学出版社,2005.Xin W X, Xie J X. Modern optimum algorithms[M]. Beijing: Tsinghua University Press, 2005.